Once you have a cap multiplier working, you have very low ripple but it stays an unregulated power supply.
Then why not go all the way, adding the little it takes, to change it into a regulated power supply.
What advantage(s) a cap multiplier can have over a regulated PSU.
Then why not go all the way, adding the little it takes, to change it into a regulated power supply.
What advantage(s) a cap multiplier can have over a regulated PSU.
Last edited:
Heat and how much of it you have to dissipate in order to protect the pass transistor. Also a regulator has to have a feed voltage well above the regulator cutoff, the greater that difference between the feed and final, then also the more heat. Then there is the heat of just the load. Then there is also the power factor losses, you'll need more transformer max current.
A cap multiplier is for when you dont need regulation but you want active ripple rejection that a regulator would provide but without the regulation. A cap multiplier has less heat because it tracks the line and load variations, goes along with it for the ride so its not fighting it. Cap multipliers work best where you know the draw will be relatively stable, like filaments or class A. If the powered circuit has wide swings in current draw then regulation may be better.
But basically drastically lower dissipation handling is the main advantage in cases where you simply dont need (or want) regulation.
For example I'm experimenting with a multiplier vs regulator at this time, in my requirements with regulation and a 2A draw current vs multiplier with a 2A draw current, my pass transistor barely gets warm in the cap multiplier (80 degrees F tops with the laser thermometer with it running for hours) with no heatsink. In the regulator version the pass transistor is hot to the touch with no heatsink (over 120 F), it would blow if I ran it too long like that.
A cap multiplier is for when you dont need regulation but you want active ripple rejection that a regulator would provide but without the regulation. A cap multiplier has less heat because it tracks the line and load variations, goes along with it for the ride so its not fighting it. Cap multipliers work best where you know the draw will be relatively stable, like filaments or class A. If the powered circuit has wide swings in current draw then regulation may be better.
But basically drastically lower dissipation handling is the main advantage in cases where you simply dont need (or want) regulation.
For example I'm experimenting with a multiplier vs regulator at this time, in my requirements with regulation and a 2A draw current vs multiplier with a 2A draw current, my pass transistor barely gets warm in the cap multiplier (80 degrees F tops with the laser thermometer with it running for hours) with no heatsink. In the regulator version the pass transistor is hot to the touch with no heatsink (over 120 F), it would blow if I ran it too long like that.
Last edited:
Regulator keeps a constant voltage.
Cap multiplier reduces ripple/noise.
They do two different jobs.
Use a cap multiplier after a regulator if you want a regulated quiet supply.
Not everything requires a regulated supply.
Cap multiplier reduces ripple/noise.
They do two different jobs.
Use a cap multiplier after a regulator if you want a regulated quiet supply.
Not everything requires a regulated supply.
It should be: cap mux first, regulator after that.
A cap mux after a regulator pretty much destroys the low impedance and rock steady voltage of the regulator. A cap mux reduces ripple but there is no ripple after a reg. But a cap mux output drops with load and creates load-dependent ripple due to it highish Zout, which the reg has been taken care of before.
Jan
A cap mux after a regulator pretty much destroys the low impedance and rock steady voltage of the regulator. A cap mux reduces ripple but there is no ripple after a reg. But a cap mux output drops with load and creates load-dependent ripple due to it highish Zout, which the reg has been taken care of before.
Jan
I understand the dissipation argument.
However, I am puzzled when considering the overall dissipation when powering a Class A versus a Class B amplifier.
When one is interested only about ripple and noise, it seems to me, they perform the same, no ? Or, is there one better than the other ?
I am thinking to use simulation to figure this out, but may be there is a known ready answer.
However, I am puzzled when considering the overall dissipation when powering a Class A versus a Class B amplifier.
When one is interested only about ripple and noise, it seems to me, they perform the same, no ? Or, is there one better than the other ?
I am thinking to use simulation to figure this out, but may be there is a known ready answer.
Member
Joined 2009
Paid Member
Many people have noted that a cap multiplier, when used as the last item on the rail, is not always associated with best sound - most likely for all the reasons pointed out by Jan. And Nelson has written somewhere around here about the necessity of following the cap multiplier by a good amount of real capacitors (ironic when you think about it).
Cap mux first followed by regulator ?
Don't you get the same result, if not better, going directly with a regulator ?
Is there an advantage to have twice a pass transistor ? One that reduce ripple, one that regulates and improves ripple.
Don't you get the same result, if not better, going directly with a regulator ?
Is there an advantage to have twice a pass transistor ? One that reduce ripple, one that regulates and improves ripple.
Member
Joined 2009
Paid Member
It likely depends on the details - reducing the ripple going into a regulator gives it a better starting point to do it's job. usually, the reason to use a cap multiplier is because yo only want to reduce riplpe without controlling the voltage; a cap multiplier will ride up and down with changes in the mains voltage level so it's self adjusting. But they are not perfect, if you over stress the supply you can starve the cap multiplier of current and it will pass ripple until it recovers. If you want to control the voltage then I think it's sufficient to have one good regulator with no cap multiplier.
I was basing my comments on -
https://www.edn.com/simple-circuits-reduce-regulator-noise-floor/
Cap multiplier output drops with load - but a typical application for input circuits has a constant load.
Usually a power amplifier does not require a regulated supply and this usage was not specified by the op.
Unless you mean the load fluctuates at audio frequency ie no decoupling?
I'm interested to learn as I may use a cap mux in a future project although probably only because I can as tbh I can't hear any noise with my current system.
https://www.edn.com/simple-circuits-reduce-regulator-noise-floor/
Cap multiplier output drops with load - but a typical application for input circuits has a constant load.
Usually a power amplifier does not require a regulated supply and this usage was not specified by the op.
Unless you mean the load fluctuates at audio frequency ie no decoupling?
I'm interested to learn as I may use a cap mux in a future project although probably only because I can as tbh I can't hear any noise with my current system.